194
PEGGION, VERDINI, COSA\I, A h D SCOFFOhE
TABLE I11 C H A R A C T L R I ZOF ~ TTI O HN R EL ~O W ~ S ET ~ E R G Y ~ I I Y I ~ I ICJ O V ~ F O R M A TFOR I O NRIi S = 1 275 A
PI o>
r’ro Ql ”I
0. Y l
0i Y i I
-4 Y
C’
D’
E’
11
I1 I1
II I1 84.37 63.37 127.43 151.55 77,7j 45.89 123.88 105.90
I1 66 27 104 09 100 07 221 33 340 17 55.83 124.96 99.28
co1 co. Et
t
Et
r
E, I E6.
11.42 0.16 4 76 6 51
47. 10 27 42 13 63 62. I 1 13 15 79.77 l23,36 104.30
I
c
L
t
t
10.36 -0.12 2.07 8.41
9.29 -0.21 -0.57 9.88
freedom to the statistical weights will also be discussed. The complete energy surface could be calculated for
the pentapeptide considered here because it is cyclic and it contains only eight variable dihedral angles in the backbone. For more complex molecules. it is impossible to calculate the complete energy surface. and other methods than those used here must be employed to find local minima within a reasonable energy range of the global minimum. The complete energy surface obtained in the present calculation can now be used to test the efficiency of such methods.2g:1
(2L)kl) N O l E ADDEUI \ PROOF. A I’ n e r h a s r,iiscd tlie point tlidt the introduction of the possibilit) for flexibility of bond ‘rngles a n d bond lengths \rould gre:it!) expand the doiiiaiii of t h e + I . $ >space n i t h i t i Lshich ring closure could be efccted. This correct. e\en though w e h a \ e no quantitative information a s to h o u the domain \ \ o d d be expanded. As w a s ciiscusscd elsen hcrc,j tlie conformation of a flexible molecu!? can be treated by a tiyo-step procedure, the first srzp being the minimization of the confortnationa! energy of thl: rigid mo!ecule h a \ ing t i x e d 15
standard bond lengths and bond angles. u i t h the second one being the energ) minimizatioii in a larger space corresponding to thc‘ flexibility of bond lengths a n d bond iiiigles s r a r f i , q from the conformation obtained bc the first stop. I n this p a p x , \\e are concerned \\it11 thc first step.
Conformational Studies on Polypeptides. Circular Dichroism Properties of Random Copolymers of Lysine and Phenylalanine in Aqueous Solutions at Various pH Values E. Peggion, A. S. Verdini, A. Cosani, and E. Scoffone Insrrrure of Otgmzrc Cliemi~rri, Uniiersrri of P o d ) i u ,35100 Pndoia, I t r i l i . Receir ed Nor riiiher 17, 1969
ABSTRACT: Circular dichroism (CD) measurements have been perforined on random copolymers of Iqsine and KCI at various p H values. On increasing the pH. copolymers containing 3 : 1 and 1 : I phen)lalanine i n 0. I molar ratios of lysine to phenylalanine undergo a conformational transition fr0.n a random coil to an ordered structure. The shape of the CD patterns. which are affexed by t h e p r e s e x e of side-chain aromatic chromophores. suggest that the ordered forms do correspond to 3 struztures. In the limits of the examined copolymer compositions. the higher is the phenylalanine content of the copolymers. the higher is the stabilitj of the 3 form in alkaline aqueous solutions.
I
t is well known that the overall stability of any conformation of natural and synthetic polypeptides is due to contributions from different forces like interor intramolecular hydrogen bonds, hydrophobic bonds, van der Waals interactions, dipole-dipole interactions, and so on.’ The relative importance of such forces are strongly dependent o n the nature of the solvent. Thus, in aqueous solutions, hydrophobic bonds among nonpolar side chains are mostly responsible for the tendency of peptides and proteins to assume definite conformations. In organic solvents like 1,2-dichIoroethane, chloroform, etc., hydrogen bonds are suggested to play a n ( I ) For a comprehciisive r e v i w on this topic, see “Structure and Stability of Biological Macromolecules,” G. N. Timashef mid G. D. Fasiiian, Ed., Marcel Dekker, Inc., New York, N. Y., 1969. ( 2 ) H. A. Scheraga in “The Proteins,” Vol. 1 , 2nd ed: H. Neurath, Ed., Academic Press, New York, N. Y . , 1963.
important role in stabilizing ordered conformations. Strong interacting organic solvents like dichloroacetic or trifluoroacetic acids compete successfully with amide groups of peptides in the formation of hydrogen bonds, and then in such media occurrence of’ordered structures is generally prevented. In a previous investigation3we studied the conforniational properties of randoni copolymers of W-carbobenzoxy +lysine (Z-Lys) and L-phenylalanine (Phe) in tetrahydrofuran as the solvent; it was found that increasing proportions of aromatic residues apparently d o not perturb the a-helical conformation of poly-“carbobenzoxy-L-lysine (PCBL). However the presence of aromatic residues in the copolymer weakened the conformational stability of the a-helical form, owing to steric interference between side chains and peptide (3) E. Peggion, A . S. Verdini, A. Cosani, m d E. Scotibne, .Mcrcroniolecrrles, 2, 170 (1969).
Vol. 3. N o . 2 , Mlircll-April 1970
RANDOM COPOLYMERS OF
backbone. i , 4 In aqueous solutions one could expect that bulky nonpolar side chains may participate in favorable hydrophobic interactions which are capable of enhancing the stability of the helical form. In the present paper we report the conformational properties of rantiom copolymers of lysine a n d phenylalanine in aqueous solutions, by CD measurements of solutions at direrent pH values. Experimental Section Solvent and Materials. Dimethylformamide (DMF) and tetrahydrofuran (7°F) were both of reagent grade and were purified as previoiusly described.j Ethyl ether (Carlo Erba R P ) has dried overnight over sodium metal and then fractionally distilled. Random Copo1)mers of Lysine and Phenylalanine. Random copolymers of lysine and phenylalanine have been prepared by decarbobenzoxylation of the parent Z-Lqs and phenylalanine copolymers. The blocked copolymers containint: different proportions of phenylalanine have been prepared by copolymerization of the corresponding Ncarboxyanhydrides (NCA’s). as described in our previous work? Copolymers 1. 2, and 3 of ref 3, containing. respectively. 3 : 1, 1 : 1. and 1 :? molar ratios of Z-Lys to phenylalanine. were decarbobenzox4lated according to the method of Fasmdn. (’2 The detailed procedure for copolymer 1 was the following. To a solution of the protected copolymer i n anhydrous: freshly distilled dioxane (1.3 g in 130 mi). chloroform (320 ml) was added. Dry gaseous HCI was bubbled for 0.5 hr through the solution which remained perfectly clear. Then dry gaseous HBr was bubbled through for 2 hr, and the deblocked copolymer precipitated as a white powder. The mixture was left standing for 2 hr. The copolymer was then recovered by filtration and exhaustively washed with dry acetone (,Merck puriss). The drq white powder so obtained was dissolved in dilute HCI and dialyzed for 4 clays against 0.01 N HCI. Dialysis was performed with a 4465-A2 dialyzing tubing (A. Thomas Co.. Philadelphia, Pa.) which retains materials with molecular weight 12.000 and higher. Finally the copolymer hydrochloride was recovered by lyophilization and dried under vacuum over phosphorus pentoxide. For copolymers 2 and 3 the same above procedure was followed. The clnly difference was in copolymer 3 which does not dissolve in aqueous HCI; in this case we dialyzed the heterogeneous copolymer solution against 0.01 N HCI. The composition of each deblocked copolymer was checked by amino acid analysis, using a Carlo Erba Model 3A27 automatic amino acid analyzer. The results of this determination art: collected in Table I. On the average. all
TABLE1
-__Lqs/Phe ratio--Copolymer
Theoretical
Found
1 2 3
3:l I :1 1 :?
3:1.05 1 : I .05 1 :3.01
samples contained about 10% moisture. The exact title of each copolymer was determined either by elemental analysis or hy micro-Kjeldhal nitrogen determinations on standard (4) H. E. Auer and P. Doty, Biocheniistrj., 5, 1716 (1966). ( 5 ) A. Cosani, E:. Peggion, A. S . Verdini, a n d M. Terbojevich, Biopolj,mers, 6, 96:; (1968). (6) G. D. Fasnian, M. Idelsoii, and E. R. Blout. J . Anirr. Chrni. Sac., 83. 709 (1961).
LYSISE 195
solutions of the various copol)mers. and b j amino acid analysis. As discussed in the previous paper.:< in all copolymers there is a statistical distribution of lysine and phenylalanine residues along the copolymer chain. Apparatus and lleasurements. C D measurements have been carried out on a Roussel-Jouan IS5 Model I I dichrograph. L\hich records directly the so-called dichroic optical density A I . - AI^. namely. the difference in absorbance between left-handed and right-handed circularly polarized radiation. From this quaiitit), the molar circular dichroism. le.and the molar ellipticity. [e].can be calculated according to the \\ell-known relations’ A I . - Ait = (AE)cc/
[e]=
3300(A€)
where c is the molar concentration and tl the optical path length. In this paper n’e report the CD curves directly recorded by the instrument. The pH of each copolymer solution was determined using a Metrohm Model E 388 precision potentiometer with a Metrohm U X combined glass electrode. CD measurements on copolymer solutions at various pH values were carried out i n the following way. The pH of each solution (-20 ml) was adjusted to the desired value by adding to the copolymers solution in 0.1 M KCI-0.8 ili NaOH with a Methrom precision microburet. Then 2.5 ml of solution was quickly transferred into the CD fused quartz optical cell and the C D spectrum was immediately run. All successive measurements were carried out on aliquot5 of the mother solution brought to the desired pH value. It was not possible to carry on measurements on copolymer 3 because the sample did not dissolve in aqueous HCI. CD measurements on copolymers I and 2 at pH higher than 10.50 were prevented by copolymer precipitation. Results
The CD spectra of copolymer 1 (Lys: Phe 3: 1) in 0.1 M KC1 solution a t various pH values are reported in Figures I and 2 . A t pH 8.68 the CD pattern is typical for polypeptides in a random coil conformation.‘ A positive dichroic band (Ae = 1.1) is located a t 218 mp, followed by a strong negative band centered belou 200 mp. The intensity of the positive band a t 218 mp is quite high with respect t o that observed for randomlq coiled polypeptides. This band, however, is quite sensitive t o the solventg~ln a n d t o the nature of amino acid residues in the peptide chain.“ In this particular case also contributions from the side chain aromatic chromophores are probably involved. At pH higher than 9.28 the CD spectrum of the copolymer drastically changes (Figure 2), and a strong negative band centered at about 218 mp becomes more and more evident. The change of the CD pattern corresponds t o a conformational transition which is strongly cooperative in character. This is even more evident in Figure 3 where the molar dichroic absorption (7) L. Velluz, G. Legrand, and M . Grosjean, “Optical Circular Dichroism,” Academic Press, N e w York, N. Y., 1965, p 62. (8) G . Holzwarth and P. Doty, J . Ai77er. Chem. Soc.. 87, 218 ( 1965). (9) .I.Steigtnaii, E. Peggion. aiicl A. Cosani, ihid., 91, 1822 ( 1969). (10) J. Steigman, E. Peggion, and A. Cosani, ihid., submitted for publication. (11) E. Peggion, L. Strasorier, and A. Cosani. ihid.. 92, 3 8 1 (1970).
196 PEGGION, VERDINI,COSANI,AND SCOFFONE
Figure 1. CD spectrum of copolymer 1 at pH 8.68. The copolymer concentration was 0.2760 g/l. in a 0.05 cm path length cell. at 218 mM is reported as a function of the pH. The conformational transition occurs within less than one p H unit; at p H 10.05 the transition is essentially complete. The complete CD spectrum at p H 10.05 recorded in the range 250-190 mp is shown in Figure 4. I n addition t o the negative band at 218 mp there is a strong positive band (Ac = 6.9) centered at 195 mp. This pattern closely resembles that of poly-L-lysine (PLL) in the p conformation,12-14 the negative band being due to the n--A* transition and the positive one arising from excitation resonance interactions of the NV1 transition. 1 5 8 1 6 The intensity of the negative band at 218 mp is about 50z lower than that found in poly-L-lysine in the @
\'Ac
I
3,1
- 1 0 1 ; 1
200
210
220 230 240
250
Mncromolecides
Figure 3. pH induced coil-8 form transition of copolymer 1 . At values at 218 mp plotted cs. pH. form. However, as pointed out by Quadrifoglio and UrryI7 in their investigation on poly-L-serine, the ptype CD patterns are variable as to position and intensity of bands when comparing different polypeptides. Moreover, in the specific case of lysine-phenylalanine random copolymers possible contributions t o the optical activity from the side chain aromatic chromophores could distort the /3 form CD pattern. From all above considerations we interpret the results of Figure 3 as a coil-to-@-formtransition. C D measurements on copolymer 2 (Lys-Phe 1 :1) in 0.1 M KCI at various p H values are shown in Figure 5 and Figure 7. At pH 4.38 (Figure 5 ) the CD pattern is quite unusual; there is a negative band (A€ = -0.32), located at 232 mp, followed by a very weak positive band centered at about 220 mp. Below 220 mp the C D pattern is quite indefinite because of the unfavorable A€/€ ratio. There is in fact a surprisingly weak dichroism in this region. The spectrum is not consistent with that of PLL and P L P in the random coil conformation. In fact both homopolymers show typical CD spectra with a positive band at 220 my (which is quite strong in PLP) and a strong negative band below 200 my. O n the contrary, Figure 5 shows that copolymer 2 exhibits very weak dichroism below 220 my. The complete CD pattern in the range 250-190 mp seems consistent with the simultaneous presence of random coil and 0 conformations; this hypothesis could explain the weak dichroism around 200 mp, which should arise from overlapping of coiled form and p structure CD bands, which are opposite in sign. The minimum at 232 mp could be
(mr> Figure 2. CD spectra of copolymer 1 in 0.1 M KC1 at various pH values: curve 1, pH 8.68; curve 2, pH 9.01; curve 3: pH 9.28; curve 4, pH 9.48; curve 5 ; pH 9.67; curve 6, pH 9.85; curve 7. pH 10.05. The copolymers concentration was 0.6822 g/l. in a 0.05 cm path length cell. (12) B. Davidson, N. Tooney, and G. D. Fasman, Biochem. Biophys. Res. Commim., 23, 156 (1966). (13) P. K. Sarkar and P. Doty, Proc. N u t . Acad. Sci. U. S., 55,981 (1966). (14) R. Townend, T. F. Kumosinski, S . N. TimasheF, G. D . Fasman, and B. Davidson, Biochem. Biophys. Res. Commun., 23, 163 (1966). (15) I
